The external efficiency of diversified secondary schools in Colombia

THE WORLD BANK EWD 57 Discussion Paper EDUCATION AND TRAINING SERJES Report No. EDT59 The Exfernal Efficiency of Diversified Secondary Schools in Colombia Eduardo Velez George Psacharopoulos February 1987 Education and Training Department Operations Policy Staff The views presented here are those of the author(s), and they should not be interpreted as reflecting those of the World Bank.  Discussion Paper Education and Training Series Report No. EDT59 THE EXTERNAL EFFICIENCY OF DIVERSIFIED SECONDARY SCHOOLS IN COLOMBIA Eduardo Velez George Psacharopoulos Research Division Education and Training Department February 1987 The World Bank does not accept responsibility for th- views expressed herein, which are those of the author(s) and should not be attributed to the World Bank or its affiliated organizations. The findings, interpretations, and conclusions are the results of research or analysis supported by the Bank; they do not necessarily represent official policy of the Bank. Copyright C)1987 The International Bank for Reconstruction and Development/ The World Bank  Abstract The paper provides empirical evidence on the most commonly used argument for diversifying the secondary school curriculum in developing countries, i.e., the hope that it would provide a closer fit between education and the world of work. A follow-up study of Colombian secondary school graduates of diversified and non-diversified (control) schools one to three years after graduation, does not provide any support to the expected external positive efficiency effects of such curriculum innovation.  I. Introduction There is a vast literature on the role of the school system for development. In particular, curricular diversification has flourished over the last two decades in many developing countries, mainly based on the rationale that diversified curricula would enable a more effective articulation between education and the labor market. Many different reasons have been put forward to explain both the expansion and diversification of secondary schooling, like economic,..social and political ones (see Ringer, 1979; Craig, 1981; l.eidenheimer, 1981; and Benavot, 1983). But since World War II, vocational and diversified education has become a central element in the educational policies of many developing countries based on the economic rationale. This rationale, resting on an apparently square logic, has appealed to many governments and substantial .educational investments have been oriented toward diversified secondary education as a strategy to correct an alleged mismatch between the traditional curricula and the skills required by the labor market. In other words, it has been assumed that unemployment is functional in its origins and that the provision of more "relevant" schooling would improve the fit to and absorption of graduates by the labor market. It was also assumed that the content of academic education unrealistically induced a demand for higher education, whereas what a country at a low level of development mostly needed was middle level skilled technicians. Thus, one of the justifications of diversification was to stop the one-way street from the secondary school to the university.  -2- Interestingly, most of the arguments in support of diversification have been put forward in a vacuum of empirical evidence on the subject. Most policy decisions in this respect have been based on the face value of intuitive logic, since no evidence existed to show whether diversified schooling had either failed or succeeded in its intended goals. To fill in part this vacuum, a study was conducted to test some of the intuitive assumptions that have influenced decisions to implement curriculum diversification. 1/ In this paper we report results on the after school experience of Colombian secondary school graduates, focusing on a comparison between the outcomes of the diversified schools (INEM) and conventional ones. "INEM" stands for Instituto Nacional de Educacion Media (National Institute of Secondary Education) which is a type of school where both academic and prevocational subjects are taught under the same roof. In this sense, INEM are very similar to European comprehensive schools. Students are encouraged to study one of several prevocational subjects, but at the same time to follow some academic courses. During the first two years students study prevocational subjects together with academic ones. After the initial two years, their study becomes more concentrated on a particular vocational field such as commerce or industry, 2/ while students continue to take some academic courses. During their final two years, the students get a more specialized education, e.g., students in the commercial track could focus in fields such as secretarial or accounting studies, students in the industrial track could focus in fields such as electricity or construction. In this way, it was expected that students would get a more varied education so that their range of future educational and occupational opportunities would be wider.  -3- On the other hand, conventional (non-INEM) schools are single track programs where the student from the first year of secondary school is exposed to an academic curriculum (colegios y liceos), or to a vocational curriculum such as agriculture, commerce, industrial arts, social services or pedagogy (univocacionales). The fist type accounts for almost 70 percent of secondary school enrollment, the second for 25 percent, while the INEM students represent less than 5 percent of the entire secondary school population.  -4- II. The Data From a methodological point of view it was decided to use a longitudinal panel design to minimize problems of selectivity bias, and because it would more easily permit to test the hypotheses related to equity and efficiency underlying the rationale of diversification. Thus, a nationwide random sample of 8051 students was used to raise baseline information in 1981, when the cohort was in the final (sixth) year of secondary schooling. 3/ In 1982, 4991 of the same students were contacted to assess their experience after they left school. Table 1 shows the sample distribution by school type and curriculum. It should be pointed out that the students from the INEM sample represent 67 percent of the universe, and that the non-INEM control sample constitutes 5 percent of its universe. Table 1: Sample Distribution by School Type and Curriculum (Number of Students/Graduates) School Type/ Baseline Follow up Curriculum (1981) (1982) INEM Academic 871 514 Agricultural 174 78 Commercial 957 588 Industrial 869 474 Social Services 380 225 Non-INEM Academic 1471 934 Agricultural 370 243 Commercial 973 557 Industrial 945 695 Social Services 274 124 Pedagogical 767 559 Total 8051 4991  -5- According to the baseline sample, INEM students (relative to non-INEM) come from families with more blue-collar heads, lower household income and less educated parents. 4/ Other indicators also point out that INEM students come from poorer families in the sense that they own less dwellings, automobiles and businesses relative to non-INEM students. It is clear that INEM schools have recruited students from less advantaged social origins, although the differences are modest. Also, the distribution of background indicators for the INEM students show a smaller dispersion relative to non-INEM students, the latter being more representative of the national distribution. As a result, one cannot conclude that diversification has broadened the school accessibility to the poor in a different way than any other type of school expansion would have done.  -6- III. External Efficiency Since the follow-up survey, one year after graduation, may be a relatively short time to reach definitive conclusions on the external effects of different types of secondary schooling on the further educational and occupational performance of graduates, what follows must be treated as indicative as to whether postschool outcomes differ according to various curriculum offering at INEM and non-INEM schools. "External efficiency" in this paper refers to the most commonly cited justification of diversification: 5/ First, diversification would diminish the demand for higher education and as a result would increase the motivation of the graduate to seek work at the end of secondary schooling. Second, diversification would lead to employment in the field of vocational specialization. Third, and by the same token, it would lead to further study by the graduate in the field of vocational specialization. Fourth, it would increase labor productivity, which may be reflected in graduate earnings. One year after graduation there were no overall differences among students from INEM and non-INEM schools in relation to their activities. About 30 percent were working full time, 37 percent were studying full time, 11 percent were studying and working simultaneously, while the rest were either searching for a job (12 percent) or were completely inactive (10 percent). However, when postgraduate activity is broken down by curriculum some differences appear. (See Table 2).  -7- Table 2: Post-school Activity by School Type and Curriculum (Percent) Postschool Activity School Type/ Study Looking Curriculum Study Work and Work for Work Inactive INEM Academic 34 30 9 14 13 Agricultural 29 27 15 24 5 Commercial 39 29 10 13 9 Industrial 36 31 8 13 12 Social Services 43 26 9 15 8 Average 37 29 10 13 11 Non-INEM Academic 38 29 13 11 9 Agricultural 34 36 11 9 10 Commercial 36 32 9 11 12 Industrial 34 32 11 13 10 Social Services 46 26 12 12 4 Pedagogical 41 29 10 12 8 Average 37 30 11 12 10 Overall 37 30 11 12 10 Academic students from non-INEM schools seem to be more inclinedto study full time or in combination with work than INEM students. Non-INEM commercial students work full time in greater proportions thantheir INEM colleagues. The INEM type of school has not increased thepropensity of academic students to enter into the labor market; only 39 percent were working, fulltime or parttime, while 42 percent of the non-INEM academic students were working in 1982. It is of interest to compare the students' expectations for postschool activities while still in school, to their actual destination  -8- one year later. Table 3 shows that there are substantial differences between what a student expects he will be doing and what he actually does. The most clear difference is related to the desire to combine study and work. While 73 percent of the students were expecting to combine these two activities, only 11 percent in fact were doing so in 1982. Table 3: Planned (1981) versus Actual (1982) Activities of Graduates by School Type (Percent) Activity INEM non-INEM Overall Plan Actual Plan Actual Plan Actual Study 18.0 37.0 26.0 37.0 23.0 37.0 Study and Work 77.0 10.0 70.0 11.0 73.0 11.0 Work 4.5 29.0 3.5 31.0 4.0 30.0 other 0.5 24.0 0.5 20.5 0.3 22.0 Although 23 percent were planning to study exclusively, 37 percent in both types of schools were actually studying a year later. In relation to work as a full-time activity, there is also a big difference between what the students planned and what they are really doing a year later -- while 4 percent were planning to work, 30 percent were doing so one year later. In summary, in a country like Colombia, the possibility to work and study at the same time is quite low and as a result graduates had to do one or the other. Nevertheless, unemployment is much higher than students expected -- only 0.5 percent expected to be inactive although 22 percent ended up inactive-in 1982.  -9- Educational Attainment Among those in further education in 1982, there are differences according to what and where the graduates had studied. Table 4 shows the distribution of full-time students by type of postsecondary institution attended in 1982 by previous school type and curriculum (contrasting academic students to vocational ones, the latter including those from the agricultural, industrial, commercial and social services tracks). Table 4: Type of Postsecondary Institution Attended by Full Time Students in 1982 (Percent) School Type Postsecondary Institution Curriculum University. Non-university Non-accredited INEM Academic 67 13 20 Vocational 49 23 28 Average 55 20 25 Non-INEM Academic 61 19 20 Vocational 59 21 20 Average 60 20 20 Overall 58 20 22 Fifty-eight percent of full-time students are in universities, and the others are split between nonuniversity and non-accredited institutions. 6/ As expected, academic students are more prone to be in a regular university, and although by a small, non-significant margin, INEM students tend to be in less proportions than non-INEM students in those institutions. The exception are academic INEM graduates who are the ones  - 10 - with the highest likelihood of being full-time students at the university level. This finding points out that diversification has not changed the propensity for postsecondary education - it has increased it among the academic students. So far the results show that diversified-education had no impact on the students postsecondary activities, and that it does not lead to less private demand for postsecondary education. But we have to isolate the effect of school type and curriculum from the influence of other factors that may also have an effect on the probability of being a full-time student. Discriminant analysis was used to see what factors have an independent effect on the probability to study, full or parttime immediately after secondary school. As shown in Table 5, several out of school as well as school factors have a significant effect. Among the first are age, sex and father's education. Younger graduates, women and graduates whose fathers have more education have a higher likelihood of being in some form of postsecondary study. With respect to school factors, Table 5 shows that the type of school does not have an independent effect - it is the curriculum that explains who is studying after school: The social services and academic students tend to have a higher probability of studying, while the industrial and agricultural students have a lower one. Those students with a high score on the academic test as well as on the ICFES test, 7/ as expected, have a higher probability to continue education beyond secondary  - 11 - school. Finally, the higher the number of schools attended during secondary schooling, the lower the probability of being a student after secondary school. Table 5: Discriminant Analysis: Studying Full or Part-Time Immediately After Secondary School Variables in Standardized Average order of Discri- Discriminant Studying Not studying minant Weight Coefficient Age .55 17.89 18.67 Academic curriculum .43 0.24 0.17 Female .37 1.56 1.48 Social services curric. .36 0.10 0.07 Industrial curriculum .36 0.20 0.22 Father's education .34 3.43 2.99 Agricultural curriculum .29 .07 .08 Academic test score .25 59.60 55.10 National ICFES test score .23 49.70 48.67 No of schools attended .19 1.26 1.34 Canonical correlation .33 Percentage correctly classified 60.1 - The model in Table 5 lists only the significant variables related to further study. Others like family income, the nature of the school (private versus public), the student's urban-rural background, the type of school (INEM versus non-INEM), school expenditure per student, teacher-student ratio, percent of qualified teachers, and an overall index of school quality did not enter in the discriminant analysis as such factors did not have an independent effect on further study.  - 12 - Table 6 shows the same model for those who were studying full or part time one year after school. Age and father's education are now the two most important factors and, consistently with the first model, diversification does not have any effect. Table 6: Discriminant Analysis: Studying Full or Part-Time One Year After Secondary School Variables in Standardized Average order of Discri- Discriminant Studying Not Studying minant Weight Coefficient Age .60 17.9 18.7 Fathers's education .33 3.43 2.99 National ICFES test score .30 49.7 48.6 Social services curric. .30 0.09 0.07 Academic test score .25 59.6 55.1 School cost per student .19 1.26 1.31 Academic curriculum .17 0.23 0.17 No of schools attended .16 1.27 1.34 Commercial curriculum .12 0.25 0.26 Canonical correlation .32 Percentage correctly classified 60.0 These results corroborate the conclusion that diversification does not reduce the demand for higher education. The curriculum has a more important effect in the sense that academic students tend to continue higher education while industrial, agricultural and commercial students do not, once they leave secondary school.  - 13 - The Relevance of Further Studies There is however the possibility that diversification leads to a closer correspondence between secondary school curriculum and the field of postsecondary studies. Table 7 shows that there is "some" curriculum correspondence regardless of school type, in the sense that just over 40 percent of commercial and industrial graduates continue to be enrolled in their respective fields, and one third of pedagogical graduates do so. Academic students, as expected, present the widest dispersion and a significant number is found in all field of studies, mainly in economics and business administration among others. However, only 20 percent of the agricultural students later specialize in agronomy; more than half of these students are studying fields that are not apparently related with their secondary curriculum, especially engineering and business administration. The latter is the field that attracts most students from any curriculum. Finally, social services graduates also exhibit, similar to the academic ones, a wide spectrum of choices. Most of them are studying business- economics courses and in second place education studies, the one nearest related to their previous curriculum.  - 14 - Table 7: Field of Post-Secondary Education by Secondary Sfol Qirriculun Post-Secoaxxary Secondary Scdool Curriculum Field of Study Academic Agricultural Crcial Imdustrial Pedagogical Social Services Education 11.6 12.2 9.0 7.9 35.1 18.1 Eine Arts 2.9 0.0 2.7 5.0 2.1 4.4 Social Science 2.0 2.0 6.1 2.8 4.3 6.9 Business Aimin- 29.5 22.4 41.1 12.6 23.9 34.4 Econatics Law 5.5 0.0 4.9 2.2 .4.8 3.8 Mathematics 5.2 0.0 7.8 6.0 4.8 3.8 Health Sciences 12.1 4.1 5.6 3.1 6.4 9.4 Enginerirg 8.8 26.5 6.8 42.8 4.8 3.8 Architecture 1.4 0.0 0.7 5.3 0.5 0.0 Agronomy 3.0 20.4 0.7 0.3 1.6 0.6 Others 17.0 12.2 14.4 11.9 11.2 16.3 Note: Nunbers are colum percentages. Recapitulating, there is a rather loose fit between the secondary school curriculum and postsecondary field of study, in the sense that it is the exception rather than the rule that those from any given secondary school curriculum later follow the same field of study.  - 15 - In Table 8 postsecondary school studies are further disaggregatE by type of school to see if diversification had an effect on the match in question. 8/ Regarding academic graduates there is a higher propensity among those from INEM to study education relative to non-INEM students. And INI commercial graduates are more prone, relAtive to those from non-INEM to Table 8: Field of Pbstsecondary Education by Type of School and Curriculum Secodary School Gurriculum Field of Academic Carmercial Social Services Industrial Agricultural Post-Secodary IEM non-INEM 4M non-INEM INEM xrn-INEM INEM norn-INEM INEM non-INEM Study Education 15.2 9.8 5.3 12.2 15.9 22.6 12.5 4.4 13.8 10.0 Fine Arts 3.7 2.5 2.1 3.2 3.7 5.7 6.6 3.8 0.0 0.0 Social Sciences 2.0 3.3 5.8 6.3 7.5 5.7 4.4 1.6 3.4 0.0 Business Admin- .23.4 32.6 54.5 29.7 37.4 28.3 9.6 14.8 17.2 30.0 Economics Law 4.5 6.0 2.6 6.8 3.7 3.8 3.7 1.1 0.0 0.0 Mathematics 5.7 5.0 8.5 7.2 2.8 1.9 5.9 6.0 0.0 0.0 Health 9.0 13.7 2.1 8.6 7.5 13.2 3.7 2.7 3.4 5.0 Sciences Enginerirg 10.2 8.1 4.2 9.0 3.7 3.8 34.6 48.9 41.5 5.0 Architecture 1.6 1.2 0.0 1.4 0.0 0.0 5.9 4.9 0.0 0.0 Agron3my 4.5 2.3 0.0 1.4 0.0 1.9 0.7 0.0 13.8 30.0 Others 20.2 15.5 14.9 14.2 17.8 13.1 12.4 11 .7 6.9 20.0 Note: Nuibers are column percentages.  - 16 - follow postsecondary studies in a field related to their previous curriculum. But the contrary is observed regarding industrial and agricultural graduates, in the sense that graduates from non-INEM are more prone to follow up their secondary school specialization. In summary, then, diversification of the curriculum has not been associated with a closer correspondence between secondary and postsecondary field of study, although the situation in this respect of the single track schools is not that better. The type of the curriculum may have a stronger impact in explaining what graduates study after secondary school. As shown in Table 9, industrial and commercial graduates (regardless of INEM or non-INEM origin) have a higher probability of continuing studying a more relevant postsecondary field of study (in relation. to their secondary curriculum), while graduates of diversified schools have a lower one. Table 9: Discriminant Analysis: Studying a Post-Secondary Field of Study Matching the Secondary School Curriculum Variables in order Standardized Average of Discriminant Discriminant Matching Non-matching Weight Coefficient Industrial curriculum .87 .36 .15 Commercial curriculum .82 .37 .25 INEM School .69 .48 .63 National ICFES test score .49 49.73 49.38 Student-teacher ratio .46 17.95 18.80 Public school .33 .80 .79 Family income .30 24141.03 28565.80 No. of schools attended .14 .13 1.28 Canonical correlation .42 Percentage correctly classified: 58.0  - 17 - Thus, in general, the results that have been discussed so far do not support the assumptions upon which diversification was adopted -- at least in relation with its implication for postsecondary education. But let us examine some evidence related to the assumed differential influence that INEM may have had, over non-INEM schools, on labor force participation. Occupational Attainment The process of occupational attainment now seems to be rather well understood in advanced contries (see Blau and Duncan 1967 for the classic work on the American case). Basically, parental socioeconomic status and intellectual ability affect the level of formal education a person receives, which in turn influences the occupational and income attainment of the person. But how does diversified secondary education in a developing country affect a person's working experience? Its advocates argue that it would lead to employment in the field of specialization. In our sample it is found that the expected occupational relevance is mixed regarding the type of school and the curriculum (Table 10). As expected, the academic students show a high degree of indecisiveness, but so is noticeable in the case of social services and agricultural students. However, this is not the case of industrial and commercial students, among whom a considerable proportion is expecting to have a job related with their specialization. Finally, the case of the pedagogical students is the one with the closest fit between curriculum and expected occupation.  - 18 - Table 10: Students Working Expectations by Curriculum and Type of School (Percentage) Curriculum Related a/ Unrelated Occupation Occupation Undecided INEM non-INEM INEM non-INEM INEM non-INEM Academic 61 60 14 15 25 25 Commercial 87 82 5 8 8 10 Industrial 67 63 25 24 8 13 Social Services 20 24 49 46 31 30 Agricultural 48 52 31 24 21 24 Pedagogical - 91 - 5 - 4 Average b/ 57 56 25 24 18 20 Notes: a/ "Related occupation" is defined, for example, as an industrial student contemplating to work as a metal worker and not as a bookkeeper; a social service student expecting to work as a social worker and not as a laboratory assistant. b/ Average excludes pedagogy.  - 19 - Of course, ex ante attitude is one thing, while actual behavior is another. So Table 11 reports what the graduates were in fact doing after they left school. Academic graduates are mainly employed as sales persons, and those from the INEM are working in higher proportion than non-INEM's in blue collar occupations such as operators and machinists. However, the INEM dimension has not helped in placing commercial, industrial and agricultural graduates in their respective occupations. Graduates from non-INEM schools in such fields are much better placed in the corresponding occupation.  - 20 - Table 11: Occupation after Secondary School by Curriculun and Type of School Academic Comercial Social Services Industrial Agricultural Occation IN4 nxn-W{ INEM rtxn-INE24 INEM rrxn-INEM INEN rxn-IE4 IME nxr-IEM Ergixerire- 0.5 0.9 0.0 0.0 0.0 0.0 2.2 5.6 0.0 0.0 related Health-related 1.0 1.9 0.0 0.4 2.0 0i0 0.0 0.4 4.3 0.0 Teading 1.0 4.4 0.9 3.0 11.3 22.9 1.8 2.0 4.3 8.5 Clerical 8.1 11.7 9.1 6.0 9.3 5.7 10.7 8.0 13.0 10.6 Secretarial 17.1 26.6 35.8 46.8 17.5 25.7 10.0 6.4 4.3 4.3 Accomting 4.0 11.1 18.8 17.9 5.2 20.0 4.8 2.8 0.0 0.0 Salesperson 32.3 22.5 21.1 20.4 33.0 17.1 17.0 14.4 41.3 19.1 Agriculture- 1.5 1.9 0.6 0.4 0.0 0.0 1.5 2.4- 8.7 36.2 related Operator, 26.2 14.6 9.1 4.2 6.1 5.8 47.9 54.4 19.6 17.0 blue-coll.ar Others 7.7 4.4 4.7 0.8 15.6 2.8 4.0 3.6 4.5 4.3 Note: Numbers are colum percentages.  -21 - The most striking finding is the low occupational match among INEM agricultural graduates. In fact, when we asked the unemployed what job they were looking for, only 10.5 percent of the diversified agricultural graduates mentioned one directly related to the curriculum - almost four time less the percentage among graduates from traditional schools. Another point-that seems to be clear is that expectations are not met after school. In all cases, the correspondence between curriculum and occupation is lower after students leave secondary school. The actual labor market situation may be an important constraint graduates have to adjust to. The assumed relationship between diversification and work after secondary school has not been proven to be clear in this data set. Additional evidence on the noneffect of INEM on the world of work is provided in the answer to the question on the usefulness of secondary studies for those who were employed one year after graduation. One half of INEM graduates that were working, against 61 percent of non-INEM graduates, answered that secondary studies were very useful in their current occupation. Adding the curriculum dimension (Table 12), the ones that support the "less useful" INEM studies are industrial and agricultural graduates. It is interesting to note that academic graduates from INEM schools gave a negative answer in greater proportions relative to non-INEM, while 43 percent of the traditional schools said that the curriculum was very useful for what they were doing (their current occupation). Only 35 percent of the INEM graduates answered accordingly.  - 22 - Table 12: Perceived Usefulness of Secondary Study for Current Work by Curriculum and School Type (Percentage) Usefulness Curriculum School None Some A lot Type Academic - INEM 20.2 45.2 34.5 - non-INEM 15.9 41.0 43.1 Commercial. - INEM 8.0 23.4 68.5 - non-INEM 4.1 22.5 73.4 Social Serv. - INEM 9.1 37.5 53.4 - non-INEM 8.0 48.0 44.0 Industrial - INEM 21.7 40.7 37.6 - non-INEM 13.9 34.8 51.2 Agricultural - INEM 32.6 39.5 27.9 - non-INEM 11.6 32.6 55.8 INEM social services graduates answered more positively than their counterparts, and this is the only case that the diversified curriculum was better evaluated regarding its perceived usefulness. But industrial and agricultural graduates gave higher marks to non-INEM schools on the same criterion.  - 23 - Labor Earnings Finally, Table 13 reports the mean earnings of those employed in 1982. Neither diversification nor curriculum are associated with any significant differences in earnings in the labor market. The same is true among graduates of the 1978 cohort, followed three years later. Statistically controlling for family background and other characteristics the mean monthly earnings of INEM graduates is 10,343 pesos, against 10,526 of non-INEM. (Psacharopoulos and Zabalza, 1984, Table 7). This finding is a strong indication that diversification of the curriculum has not increased labor productivity. Table 13: Monthly Earnings by School Type and Curriculum, 1982 (in Pesos) Curriculum INEM non-INEM Academic 10,324 10,340 Agricultural 10,209 9,249 Commercial 9,487 9,616 Social Services 9,352 10,903 Industrial 9,485 9,943 Pedagogical (9,829) Average 9,723 10,010  - 24 - IV. Summary and Conclusions Evidence has been presented on the extent to which the new INEM (diversified) schools have had a differential effect on some indicators of external efficiency, as compared with conventional secondary schools. There is virtually no difference in the proportions of INEM and non-INEM students who want to continue studying, and contrary to the expectation of the policy makers, the proportion of INEM academic graduates that enter university is higher than the one from traditional academic graduates. Also, INEM graduates are not more likely to follow their secondary school curriculum in their postsecondary studies, with the exception of commercial graduates. In relation with the world of work, INEM students seem to be more congruent in their expectations regarding the relationship between curriculum and area of employment. But one year after school non-INEM graduates are more likely to be employed. INEM agricultural graduates do not mainly enter agricultural jobs, and to some extent this is also true regarding industrial and commercial graduates. In general, non-INEM graduates seem to be in occupations more closely related to their secondary school curriculum than INEM graduates. And neither diversification nor curriculum type are associated with any earnings advantage among those employed one to three years after graduation. Thus, in the case of Colombia, the evidence does not support the rationale on which diversification was justified. In fact, such curriculum innovation does not appear to be better nor worse than conventional schooling. Such result is corroborated when using data from the 1978  - 25 - cohort of secondary graduates, i.e., three years after graduation (See Psacharopoulos, Velez and Zabalza, 1986). This removes the caveat that one year may not be enough to evaluate curriculum diversification. We do not believe that the results presented here would change even if the graduates were followed up for several additional years. The point is that if diversified education has not accomplished its objectives immediately after graduation, it is unlikely to do so when many other extraneous factors intervene in the graduate's career. Perhaps if less emphasis were given to diversification as prevocational education and more emphasis given to its pedagogical value (combined with on-the-job-training for specialized skill provision), this would have led to a more efficient educational system and less frustration regarding an apparently failed reform. 9/ By the same token, emphasis on the agricultural curriculum could have been reduced, as such prevocational training could be achieved more efficiently by means of purely vocational agricultural programs in rural areas. The case of the social services curriculum could also be reviewed, as graduates from such curriculum do not have a well-defined labor market. And yet, paradoxically, there is a wide spectrum of problems in both urban and rural communities where such graduates could help. Perhaps some years of national service could have been required from such students (as well as agricultural ones) to give them the opportunity to work in related occupations and at the same time to help solve urgent community problems.  - 26 - Footnotes 1/ For more comprehensive reports on the findings of this study conducted in Colombia and Tanzania, two countries where programs on diversification have been well implemented, see Psacharopoulos and Loxley (1985); and Psacharopoulos, Velez and Zabalza (1986). 2/ INEM also offer specializations in agriculture and social services, as well as a traditional academic orientation. 3/ A representative number of high school students was obtained in proportion to their enrollment in the diffetent curricula in 128 schools located in 27 cities accross the country. For a more detailed description of the sample framework see Psacharopoulos and Loxley (1985). 4/ For the data on which conclusions in this paragraph are made, see Psacharopoulos and Loxley (1985), Chapter 5. 5/ For a wider treatment of external efficiency, including statistical analyses of the effect of the curriculum on earnings and unemployment, see Psacharopoulos and Loxley (1985), Chapter 5. 6/ A non non-accredited institution is one which does not offer degree programs. 7/ This is a test administered by the national examination board for entry into higher education. 8/ Pedagogy students are excluded because INEM, as it was already mentioned, do not offer such curriculum. 9/ In Psacharopoulos, Velez and Zabalza (1986) it is shown that INEM schools, other things being equal, are associated with a higher level of cognitive achievement.  REFERENCES Benavot, A. (1983). The Rise and Decline of Vocational Education Sociology of Education. Vol 56, 63-76. Blau, P. and Duncan, O.D., (1967). The American Occupational Structure New York: Wiley. Craig, J. (1981). The Expansion of Education Review of Research in Education, 9, 151-2/3. Heidenheimer, A. (1981). Education and Social Security entitlements in Europe and America. In The Development of Welfare State in Europe and America, P. Flora and A. Heidenheimer (eds), New Brunswick: Transaction Books. Psacharopoulos, G., and Zabalza, A., (1984). The Effect of Diversified Schools on Employment Status and Earnings in Colombia Economics of Education Review, 3, 315-31. Psacharopoulos G., and Loxley, W. (1985). Diversified Secondary Education and Development: Evidence from Colombia and Tanzania. Baltimore: The Johns Hopkins University Press. Psacharopoulos, G., Velez, E. and Zabalza, A. (1986). Una Evaluacion de la Educacion Media en Colombia. Bogota: ENLACE. Ringer, F., (1979). Education and Society in Modern Europe. Bloomington: University of Indiana.